Clinical manifestations

Presentation and course

The complaints and findings in patients with failed back surgery syndrome depend very much on the causes of individual cases. Patients with the disorder report varying degrees of back or leg pain with or without neurologic deficit that they regard as debilitating or disabling. Symptoms are commonly present for more than 6 months. Those most severely affected have cauda equina dysfunction. Symptoms may predominate in the back (axial) or legs (radicular or pseudo radicular). Axial symptoms may include unremitting or episodic low back pain accentuated by static postures or motion. Lumbar spine instability has been implicated as a cause of axial pain and may include "overt" instability, such as that seen with dynamic radiographs and iatrogenic spondylolisthesis, or "covert" instability as described by Masferrer, in which painful micromotion is suspected at the interspace level (47). In both instances, the spine is unable to tolerate static postures or loading without pain. A disruption of the lumbar disc or "joint space" with the index procedure may be sufficient to cause these symptoms although posterior column and support ligament damage generally increase instability. Fox demonstrated radiographic instability in 30% of postoperative decompressive laminectomy patients. This figure increased dramatically to over 70% if preoperative subluxation was seen (24). Some feel that cases of failed back surgery syndrome due to instability resulting from decompression may be reduced by performing lumbar fusion with the index procedure (14). Extremity symptomatology may evidence radiculopathy (nerve root pain with demonstrable neurologic deficit) or radiculitis (nerve root pain only). Referred pain from spinal structures such as the annulus (groin pain) and facet joint and sacro-iliac joint (truncated radicular distribution) may mimic radicular complaints. The etiologies of radicular pain in failed back surgery syndrome include recurrent disc herniation, lateral or foraminal stenosis, nerve root injury, and epidural fibrosis. Foraminal stenosis may result from discectomy with interspace collapse or iatrogenic spondylolisthesis following facetectomy and laminectomy. Instability of the spine can produce episodic nerve entrapment with certain postures or activities accentuating subluxation. Epidural fibrosis with tethering of the thecal sac and nerve roots may result in local and referred pain as well as radicular symptomatology (57).

Painful internal fixation with spine fusion results from hardware malpositioning with nerve root injury or entrapment, loosening or dislodgement with soft tissue and neural element pain, and the generation of an inflammatory response with focal or referred symptoms. With hardware in place, chronic low-grade spondylodiscitis or frank infection are heightened concerns. Failed fusion can also be asymptomatic. The dissection required for internal fixation placement and the significant muscle compression generated by fixed retractor systems utilized in posterolateral fusion procedures with pedicle screw fixation has been shown by histological study and EMG to cause areas of permanent muscle dysfunction and fibrosis described as "fusion disease."

Graft harvest, which commonly takes place from the iliac crest, may involve disruption of the sacroiliac joint and injury to the cluneal nerves with focal muscle or neuropathic pain.

Arthrodesis places greater mechanical demands on the segments above and below the fusion and, depending on the number of segments involved, the sacroiliac joints. The so-called transition segment may experience an accelerated degenerative change with the potential for herniation, stenosis, or instability resulting in failed back surgery syndrome, due to symptoms at a previously untreated level. There is growing interest on differential risk of spine surgeries and relationship to adjacent level disease. A preliminary series implicates that the more rigid posterior lumbar interbody fusion results in a higher 10-year reoperation rate for adjacent level disease than posterolateral lumbar fusion. Age greater than 60 years was the only other risk factor for adjacent segment disease. This is important as the outcome after adjacent segment operation is far worse than the outcome of the original surgery (42). Flat back syndrome is an untoward sequelae of lumbar fusion; it is more common with multiple level fusions. This loss of lumbar lordosis may cause chronic low back and hip pain.

Arachnoiditis is highly variable and has the potential for radicular or cauda equina symptomatology. At times it has been associated with devastating neurologic deficit, and it may complicate diagnostic and surgical approaches to the lumbar spine. Potentially progressive, this process may mimic recurrent disc hernia, stenosis, and epidural fibrosis, with relatively early onset within weeks to months after documented surgical treatment.

Clinical vignette

A 38-year-old machinist was injured while lifting a 50-pound pail at his place of employment. He had acute onset of left-sided low back and buttock pain that, over 24 hours, progressed to include left leg pain radiating down the posterior thigh and leg. Medications and physical therapy were ineffective in managing his pain. MRI showed a left L4-5 paracentral disc herniation. Epidural injection on two occasions provided only short transient improvement. He proceeded to surgical discectomy that improved his severe sharp left leg symptoms but with residual left leg and back pain. Physical therapy and work hardening resulted in an exacerbation of his severe back pain. MRI 5 months after surgery showed epidural scarring around the left L5 nerve root only whereas a study 14 months after surgery showed epidural scarring and a small left L4-5 posterolateral disc protrusion. EMG 18 months after surgery showed an old left L5 radiculopathy. Follow-up transforaminal epidural injections at L4-5 and L5-S1 resulted in no transient or sustained improvement. Repeat MRI 20 months after surgery showed a large recurrent left posterolateral L4-5 disc herniation. Repeat discectomy was completed 22 months after his first surgery. Severe pain in the left leg was resolved, but he had persistent low back pain and activity intolerance with increased left leg and back pain with any activity that did not improve with physical therapy or work hardening. Follow-up MRI scans show only left L4-5 epidural fibrosis and disc degeneration with no significant disc protrusion or herniation.

Biological basis

Etiology and pathogenesis

Although the potential causes are numerous, clinical studies have shown that some etiologies account for a greater proportion of failed back surgery syndrome cases. Approximately 95% of patients can be provided a specific diagnosis. There is approximately an equal distribution between the incidences of nonsurgical and surgical diagnoses (64). Waguespack presented a study of 183 consecutive patients with failed back surgery syndrome, in which a predominant diagnosis was identified in 95% of the cases. She found foraminal stenosis in nearly 30%, painful lumbar disc disorder with or without stenosis or previous fusion in 24%, pseudoarthrosis in 15%, neuropathic pain in 9%, recurrent disc herniation with or without stenosis in approximately 8%, iatrogenic instability in 5%, a psychological basis in 3%, infection in 2%, and arachnoiditis in 1% (70). Similar findings were seen in another study where the causes of failed back surgery syndrome were noted to be foraminal stenosis 29%, symptomatic degenerative disc disease 22%, pseudoarthrosis 14%, neuropathic pain 10%, recurrent disc herniation 12%, facet joint pain 3% and sacroiliac joint pain 2% (61). Ha and colleagues demonstrated in a prospective cohort study a markedly increased incidence of sacroiliac joint degeneration after lumbar spine fusion (29). In a 5-year clinical review of 38 consecutive patients with failed back surgery syndrome selected for lumbar arthrodesis by the author, the most common diagnosis was recurrent disc hernia with or without foraminal stenosis, followed by painful lumbar disc disruption and iatrogenic spondylolisthesis (30). In addition to structural pathology, a psychosocial basis is recognized as a contributing factor in most cases of failed back surgery syndrome (44).

The pathogenesis of symptoms varies from case to case. Lower back pain accompanies an array of musculoskeletal disorders, spinal instability, and microtrauma with spinal movement. There can be nerve root compression by residual or recurrent disc, scar tissue, foraminal or canal stenosis, or osteophyte. Spinal instability may contribute to nerve root trauma and irritation. Inflammation, neuropathic pain, and functional overlay may contribute to symptoms. Multiple mechanisms are responsible for most cases of failed back surgery syndrome.

Epidemiology

It has been estimated that as many as 15% of patients have some degree of failed back surgery syndrome after lumbar procedures (57). This volume of patients is significant when one considers that in 1990, 279,000 lumbar operations were performed on adults in the United States. Of those procedures, 46,000 were spine fusions. Between 1979 and 1990, the rate of lumbar spine fusion for adults rose from 13 to 26 per 100,000. Of those fusions, 51% were to treat degenerative disc disease. Although various internal fixation systems were available in 1990, the approval of interbody fusion cages by the FDA to treat painful degenerative disc disorders in 1996 has resulted in a significant increase in the volume of these cases (30). The added availability of disc arthroplasty increases the complexity of this discussion. It is now estimated that 100,000 patients undergo fusion or arthroplasty for degenerative lumbar disc disease in the United States annually. With greater numbers of spine surgeons receiving advanced training and performing more complex procedures, an increase in the complexity of failed back surgery syndrome is inevitable.

Prevention

Although careful selection of patients for lumbar discectomy and meticulous surgical technique are associated with better outcome, failed back surgery syndrome cannot be prevented. In an animal model, mitomycin C applied locally effectively reduced epidural fibrosis, completely avoided dural adherence, and induced no side effects (43). The random application of elastic protein-based polymers post laminectomy may also reduce the formation of epidural fibrosis and the tethering of the underlying neural elements by the postlaminectomy epidural fibrosis (02).

Prevention includes the management of risk related to bio-psychosocial risk factors associated with failed back surgery syndrome. In a systematic review, it was noted that failed back surgery syndrome was associated with lower level of education, higher level of preoperative pain, less work satisfaction, longer duration of sick leave, higher levels of psychological complaints, and more passive avoidance coping function with anxiety and somatization (18). Some of these factors could be addressed presurgically although the impact of such interventions is unknown. Presurgical depression also impacts outcome (63). In their prospective study of 102 patients undergoing spine surgery stratifying for severity of pathology, Sinikallio and colleagues correlated Beck Depression Inventory scores and depressive burden scores with poorer self-reported symptoms, higher symptom scores, and lower walking capacity.

A retrospective cohort controlled trial of 50 patients undergoing open discectomy for disc herniation demonstrated the value of pre- and post-surgical pregabalin in improving pain and function outcomes after surgery over 6 months. The study demonstrated the positive effect increased over the 6-month study period (13).

Differential diagnosis

Associated or underlying disorders

Recurrent or continued symptoms following lumbar disc surgery is almost always related to the failure of the surgery to address the original problem, complications of the surgery itself, or recurrent or new disc herniation.

Complications include infection, arachnoiditis, pseudomeningocele, spinal instability, and epidural fibrosis.

Time course of onset of symptoms may assist in diagnosis. Continued pain right after surgery may indicate wrong or incomplete surgery, technical error, incomplete fusion, or poor psychological risk factors for surgery. Temporary relief followed by symptoms should alert to reherniation or infection as possibilities. A delay in symptoms for several months may be due to instability after a brace is removed or may be due to arachnoiditis or fibrosis. A more delayed onset of symptoms might be due to degeneration or instability adjacent to prior surgery.

In addition, expectations need to be appropriately managed. Forty percent of those who have a surgical treatment for a disc herniation achieve their premorbid level of activity, with 10% having increased back pain and 10% having increased leg pain.

Diagnostic workup

The diagnostic approach includes examination, plain and dynamic radiographs, imaging, invasive diagnostic tests, orthotic trials, and psychological profiling. The extensiveness of the evaluation is tailored to the clinical picture. MRI scan is useful in identifying recurrent lumbar disc hernia, degenerative disc disease, or disc dehydration, epidural fibrosis, peridiscal inflammatory or fibrovascular marrow change, and high intensity zones of the posterior annulus.

Reports have documented a poor correlation between early MRI findings and symptoms in failed back surgery syndrome increasing the risk of failure, when revision surgery is weighted heavily on those results (23). A clinical series suggests that additional benefit in directing treatment is seen with magnetic resonance neurography of the lumbosacral plexus in these cases, although a 3T machine was used in this non-cohort, non-case controlled series (19).

In an evidence-based treatment plan, CT-myelography, diagnostic spinal block, and discography proved useful (70).

CT-myelography is used as a complimentary study to MRI. Boney detail, particularly in multi surgery cases, is difficult to appreciate with MRI scanning. Image degradation from internal fixation also renders MRI suboptimal in instrumented fusion cases.

Further, in a relatively straight-forward surgical problem such as spinal stenosis, Riew reported that CT-myelography altered therapy in 49% of patients whose operative plans were based on MRI alone (55).

Selective nerve root block helps to determine the source of extremity pain, especially when multilevel surgery has been performed and symptoms, imaging, and examination are confusing or contradictory. The epidurogram (simultaneously performed with the block) complements data provided by MRI and CT-myelography, particularly in cases of foraminal stenosis. Also of use are blocks of facet joints and internal fixation as well as the iliac crest donor site in fusion cases.

Because there are no pathognomonic findings on examination or MRI scan that confirm a painful disc disorder, discography has been shown to be particularly useful in these cases (62).

Provocation discography may identify a painful previously operated segment, painful adjacent discs, or painful interspaces beneath posterolateral fusions. Psychology profiling may identify nonorganic pain as a predominant factor and is somewhat predictive of the response to surgical intervention (11). Many feel that abnormal scores on the hypochondriasis and hysteria scale predict poor results with further aggressive treatment in failed back surgery syndrome. Electrodiagnostic studies are rarely diagnostic in failed back surgery syndrome, suspected to be due to the large number of patients without neurologic abnormality and the inability to effectively evaluate the paraspinal muscles after recent surgery. Bracing trials are recommended by some authors when selecting patients for fusion (46; 47). Ultimately, the diagnostic evaluation should lead to a specific pain generator. This approach avoids presumptive diagnoses and structural surgical approaches, unless a surgical lesion that can be reasonably expected to account for the patient's pain has been identified.

Management

Lumbar repeat laminotomy, discectomy, or canal exploration is indicated primarily for recurrent or persistent radiculopathy with demonstrated nerve root entrapment. Ozgen reported a series of 114 patients; "true" disc hernia was found in 58 of the patients at re-exploration. He concluded that recurrent disc herniation is the most important indication for re-exploration, and this group had the best outcomes compared to those in which epidural fibrosis was identified (54).

Lumbar posterolateral fusion, with or without canal exploration and internal fixation, is considered when spinal instability has been identified either with dynamic radiographs, discography, or other selection techniques (45; 24; 31; 59; Hacker unpublished data). The literature reports variable success with arthrodesis. Markwalder reported good or excellent results for 79% of patients in whom a meticulous preoperative evaluation had concluded that failed back surgery syndrome was due to spinal instability (45). Biondi found that only 49% of patients treated with posterolateral fusion for failed back surgery syndrome reported good results and correlated poor results with Workers' Compensation cases, less than 6 months pain-free interval from the previous operation, male sex, history of psychiatric illness, and a predominant diagnosis of perineural fibrosis (07). Schnee reported that despite older age, adult patients with a single failed operation and spondylolisthesis may obtain significant benefit from surgery. Patients with at least two prior operations or preoperative pseudoarthrosis fared particularly poorly in his study (60).

Anterior lumbar interbody fusion has been reported as a safe and effective procedure for the treatment of failed back surgery syndrome. Thirty-three patients with a preoperative diagnosis of failed back surgery syndrome, with degenerative disc disease (n = 17), postsurgical spondylolisthesis (n = 13), or pseudarthrosis (n = 3), underwent anterior lumbar interbody fusion. Back pain, leg pain, and functional status improved significantly (20).

Interbody fusion cages have been seen in clinical trials to benefit patients with painful degenerative disc disorders that may be related to micromotion or other primary disc space pathology (31; 41). White and coworkers gave an initial report of their experience with threaded interbody fusion cages and posterior lumbar interbody fusion for failed back surgery syndrome due to recurrent disc hernia. Improvement in McGill pain questionnaire and Prolo scale results were seen postoperatively; however, the followup period in their series was short, with most patients less than one year after surgery (71). In the author's experience with 38 patients treated with interbody fusion cages for failed back surgery syndrome over a 5-year period, 55% found good or excellent results (31).

Long opined that the nature of failed back surgery syndrome is one in which only nerve root compression and spinal instability will respond to repeat operative intervention (44). Although the studies discussed document benefits from revision surgery, failures are common and occur with much greater frequency than with index spine procedures.

Percutaneous endoscopic transforaminal lumbar interbody fusion for high-risk surgical patients currently carries a high 36% complication rate and 26% reoperation rate, making it a procedure that cannot be recommended until significant technical improvements in the procedure are accomplished (35).

Early clinical experience with endoscopic lumbar foraminotomy is promising as it avoids a higher risk surgery treatment and provides excellent short term pain control in select patients identified by imaging and selective nerve root blocks. Patients are selected with prior surgery failure with continued foraminal stenosis (01).

Spinal cord stimulation has been a treatment option for failed back surgery syndrome for over 20 years. A MEDLINE search from 1966 to 1994 of spinal cord stimulation in the French and English literature for failed back surgery syndrome found that 59% of patients studied had at least a 50% reduction in pain. The authors determined that the lack of a control group precluded conclusions regarding the effectiveness of this approach (67). Burchiel and colleagues reported 1-year followup data from a multicenter trial of spinal cord stimulation, in which 88% of 70 patients had lower back pain and lower extremity symptoms. Although multiple pain assessment questionnaires showed that pain was successfully managed in 55% of the patients (patients reporting at least 50% pain relief and reporting the procedure beneficial), no significant improvement in medication usage or work status was noted (10). Ohnmeiss found a statistically significant improvement in lower extremity strength in painful legs after spinal cord stimulation for failed back surgery syndrome (52). There was a reduction or discontinuation of narcotic medication in 66% of the patients who were taking narcotics before spinal cord stimulation in the study. Selection of patients for spinal cord stimulation is, at best, difficult. In an effort to determine who should undergo the procedure, preoperative testing was correlated with outcome in a study of 40 patients (85% with failed back surgery syndrome) at 3 months post spinal cord stimulation. The most important predictors were patient age, Minnesota multiphasic personality inventory, depression subscale D, and the evaluative subscale of the McGill pain quotient (11). Using these data, the authors could predict outcome in 88% of patients. An industry sponsored study by North and colleagues indicates that alternative programming with subperception stimulation at 1 kHz in patients who are refractory to paresthesia-based stimulation may improve efficacy based on measures of pain and functional outcome (48). High frequency or burst stimulation may be more efficacious than standard programming in improving pain, sleep quality, and depressive symptoms in patients with predominately back pain (39). A relationship between these symptomatic changes and elevation in interleukin 10 was demonstrated in a small sample (38). A 24-month randomized controlled trial of standard spinal cord stimulation compared to 10 KHz spinal cord stimulation showed a greater than 50% reduction in pain scores favoring the high frequency stimulation group (37). Even out to 24 months, back pain responders were 76.5% versus 49.3% and leg pain responders were 65.1% versus 46%. Improvements in satisfaction and Oswestry Disability Index were noted, although changes in opioid pain medication use and return to work were not assessed. Study limitations included the inability to blind due to the different parameters of the stimulation (37). Studies indicate that evaluating spinal cord stimulation effectiveness, utilizing multiple psychosocial dimensions of well-being and participation, gives a fuller expression of the patient’s perspective of their experiences with spinal cord stimulation beyond simply the pain dimension (72).

A significantly more challenging cohort to address with spinal cord stimulation is the patient with failed back surgery syndrome with chronic back pain. A small controlled trial demonstrated the added benefit of subcutaneous stimulation to treat back pain in patients with leg pain improvement after traditional spinal cord stimulation (69). The subsequent open label arm demonstrated the added benefit of peripheral nerve stimulation in this patient population, although nearly a quarter required reoperation due to an adverse event (68). Multicolumn spinal cord stimulation demonstrated benefit in a small percentage of patients with back pain with failed back surgery syndrome compared to those treated with optimal medical management (56).

A population-based controlled cohort study evaluated the efficacy of spinal cord stimulation in worker’s compensation recipients (66). There were 51 patients seen in the pain clinic who received a spinal cord stimulation, 39 who were seen in the pain clinic who did not get spinal cord stimulation, and 68 who received usual care. Pain, function, work status, and medication use was measured at baseline and at 6, 12, and 24 months. Based on these outcome measures, less than 10% achieved success, with decrease in leg pain and opioid use being the common measures that improved. As a group, the spinal cord stimulation group had lower leg pain and higher function scores but higher rates of daily opioid use. By 12 months, there were no differences in the three groups. By 18 months, 19% had their spinal cord stimulation device removed.

A follow-up study looked at the cost effectiveness of spinal cord stimulation in this trial (34). Their analysis showed that the mean cost of care over the 24 months was $52,091, which was over $17,000 higher than the pain clinic group and $28,000 higher than the usual care group. Statistical analysis showed that there was less than 5% likelihood that spinal cord stimulation would be the most cost-effective treatment for any patient. A 12-year retrospective study of insurance data showed the overall cost effectiveness of spinal cord stimulation compared to non-spinal cord stimulation, although the nature of the other treatments prescribed and their relative clinical effectiveness were not reported. Significant between payer-type cost differences were identified (21). A systematic review shows that spinal cord stimulation is not cost effective over a 2-year horizon. It is cost effective over a longer period of time (15 years) if compared to persons who in that horizon have subsequent surgical intervention, although comparison was made versus best clinical care but not epidural injection nor adhesiolysis (50).

Early case series have generated interest in prospective trials in the use of dorsal root ganglion stimulation in failed spine surgery syndrome (36). Choosing the level of stimulation was based on identifying the pain-paresthesia overlap, especially in those with leg pain. L2 was the level of choice in patients with back pain with one to three leads placed per patient. Half of the patients screened were ruled out due to foraminal epidural fibrosis. Eleven of 13 trialed went on to implantation, with up to 80% pain improvement. Significant postsurgical complications indicate the need to maximize the technique (36).

A placebo controlled trial of transcutaneous magnetic stimulation 5 days per week for 2 weeks in 20 patients with failed back surgery syndrome showed improvement in visual analog scale, neuropathic pain score, Oswestry Disability Index, Beck Depression Inventory, and Pittsburgh Sleep Quality Index that persisted through the 3-month follow-up. The protocol consisted of primary motor cortex stimulation as per typical protocol (12).

Intrathecal morphine infusion has been used to treat malignant pain for nearly two decades. Its application to chronic nonmalignant pain is relatively new. Its success has been variable. Patients with intrathecal morphine infusion reviewed by Brown continued to show high pain levels and severe disability (09). However, patients on average described themselves as having a 64% improvement in pain and a 48% improvement in functioning. In a study of 40 patients, 47% of whom were diagnosed with failed back surgery, Anderson found significant improvement over baseline levels of visual analog scale pain at each followup examination after implantation. Fifty percent reported at least a 25% reduction in visual analog scale pain after 24 months (03). Device-related complications requiring repeat operations were experienced by 20% of the patients in that study.

A retrospective review comparing intrathecal morphine infusion and spinal stimulation for patients with lower back pain and lower extremity pain supported an approach in which infusion is used for patients with buttock, axial, and bilateral leg pain unresponsive to stimulation. The percentage of patients with failed back surgery syndrome in this study was not specified (32).

A retrospective study by Hayek and colleagues of intrathecal hydromorphone and bupivacaine showed improvements in visual analog scale (VAS) score and decreased oral opioid use (33). However, dose escalation of 133% at 6 months, 78% over the second 6 months, and 48% in the second year was seen.

In a small percentage of patients with failed back surgery syndrome, other invasive treatment options such as internal fixation removal, radiofrequency facet denervation, or epidural block may be of value. The lack of studies with data on which to base conclusions makes these treatments, at this time, anecdotal.

Ganglionectomy is an aggressive approach that should be condemned. Long-term followup for this approach suggests that an extremely low percentage of patients achieve any benefit, and many experience procedure-related neurologic deficits (49).

Adhesiolysis is theorized to be effective in failed back surgery syndrome as it is a treatment directed to one of the purported causes, namely epidural fibrosis. A systematic review stated level I or II-1 evidence for short- or long-term relief of symptoms (22); however, this was based on only three studies, two without control groups. A subsequent systematic review of endoscopic adhesiolysis included nine studies, four of which are from 2018 or later, with a meta-analysis of 392 patients (27). VAS score improvement of 3.4 at 6 months and 2.8 at 12 months was achieved, with Oswestry improvement of 19.4% and 19.8% at 6 and 12 months from a baseline of 62. Twenty-two patients had spinal fluid leaks, with six that were symptomatic with headache.

An early case series on epidural injection of sodium 2-mercaptoethanesulfonate (a mucolytic) has demonstrated improvement in pain and function and reduction in opioid medication use in patients with failed back surgery syndrome and imaging demonstration of epidural fibrosis (15).

The addition of preoperative irradiation may improve outcome in patients who undergo reoperation for recurrent radicular pain associated with a significant amount of peridural fibrosis (26).

Nonsurgical approaches to failed back surgery syndrome, including physical therapy, rehabilitation, and interventional spine therapy, are well accepted alternatives to invasive treatment. Their success in this patient population has not been demonstrated. A study utilizing maintenance opioid therapy described results for 21 patients treated for a mean of 32 months. Pain improvement was significant, but only two patients returned to work. The authors found no drug diversion, addictive behavior, or organ toxicity and stable doses (59). Tizanidine has shown a dose-dependent antinociceptive activity presumably due to decreased release of substance P and reduction in excitatory amino acid activity at the spinal level (58). Multidisciplinary pain clinic approaches are indicated in patients with psychosocial risk factors for poor outcome and have reported patient improvement with costs that are generally lower than surgical intervention. These patients generally don't have recognized structural pathology that would benefit from surgical treatment. A clinical series of 34 patients demonstrates the value of intensive inpatient multidisciplinary rehabilitation in improving pain, improving function, decreasing depression, and decreasing pain medication use in this population (04).

Outcomes

Fritsch reviewed revision surgeries performed on 136 patients, 34% of whom had multiple revisions. Initially, 80% found satisfying results, but this decreased to 22% with long-term follow-up. With multiple revision procedures, instability and epidural fibrosis were found more frequently than recurrent disc hernia, with scar predominating in 60% of those cases (25). Of concern are studies looking at longer follow-up intervals for patients with failed back surgery syndrome treated with fusion. A study showing 84% of patients achieving excellent and good results at 3 to 6 months found that only 56% maintained that outcome at 1 to 2 years (17).

Failed back surgery syndrome continues to be a problem affecting all lumbar spine procedures from simple to complex. Avoiding the creation of this condition by meticulously selecting patients for surgical intervention would conceivably have a greater impact than any treatment measure described to date (57). An evidence-based diagnosis and treatment approach that recognizes the limited etiologies that respond to surgical treatment will hopefully avoid multiple unsuccessful revision surgeries.

Special considerations

Pregnancy

Pregnancy imposes mechanical stresses on even normal backs; symptoms are probably worse in women with previous lumbar disc surgery. Such women present diagnostic challenges, and there are few guidelines for their care. If symptoms are limited to pain, treatment should begin with conservative measures, many of which are useful in pregnancy without failed back surgery syndrome (06). Progressive or severe neurologic deficit usually requires MRI scanning (no radiation exposure) and, possibly, surgery. Consideration of ultrasound-guided caudal epidural injection would be reasonable in this setting.

Anesthesia

Nerve stimulation and EMG recordings are occasionally useful in monitoring surgery; neuromuscular blockade must be avoided in these cases.